Program evaluation in the correctional sciences questions the values or efficacy of correctional programs. Program evaluation has its own language, conventions, and misconceptions which the author discusses in detail. Additionally, applied behavior analysis plays a major role in questions of program evaluation in the correctional sciences because it is a philosophy explaining why people act as they do, a system of logical inquiry, and a technology for changing behavior. Program evaluations often grow out of the development of a new or novel treatment approach. The evaluation, then, is often used to measure the efficacy of the new approach. This gives rise to many practical questions. Is the program grounded in meaningful theory? Who will carry out the evaluation? What questions will the evaluation attempt to answer? As these questions are explored, evaluators should be aware of three specific threats to the internal validity of correctional evaluations: sample selection difficulties, subject maturation, and statistical regression effects. The author points out that correctional evaluation is a potentially powerful tool that has produced less than satisfactory results to date. The unsatisfactory results occur because the questions asked by correctional evaluators have tended to be disappointing or equivocal and indefinitive. 99 references.